Cephalonia-Lefkas Transform Fault Zone (CLTFZ) Complexity: Insights from 2015 Lefkas Earthquake Sequence

Journal of Geography, Environment and Earth Science International

21(1): 1-12, 2019; Article no.JGEESI.49088 ISSN: 2454-7352

Cephalonia-lefkas Transform Fault Zone (CLTFZ) Complexity: Insights from 2015 Lefkas Earthquake Sequence

Andreas Karakonstantis1*, Kyriaki Pavlou1, Vasilis Kapetanidis1 and Georgios Bozionelos1

1National and Kapodistrian University of Athens, Ilissia, Zografou 157 84,Grece.

Authors’ contributions

This work was carried out in collaboration among all authors. Author AK designed the study, managed the literature searches, wrote the protocol and the first draft of the manuscript. Author KP performed the Coulomb Stress Transfer Analysis, author VK managed the catalogue unification, analysis and statistics and author GB determined the Focal mechanism solutions. All authors read and approved the final manuscript.

Article Information

DOI: 10.9734/JGEESI/2019/v21i130119 Editor(s): (1) Dr. Kingsley Eghonghon Ukhurebor, Lecturer, Department of Physics, Edo University Iyamho, Edo State, Nigeria. Reviewers: (1) Agu Eensaar, Tallinn University, Estonia. (2) Orhan Polat, Dokuz Eylul University, Turkey. (3) Vinay Kumar, Florida State University, USA. Complete Peer review History: http://www.sdiarticle3.com/review-history/49088

Received 06 March 2019 Accepted 13 May 2019 Original Research Article Published 18 May 2019

ABSTRACT

In order to define a better model for the Cephalonia-Lefkas Transform Fault Zone the sequence of 2015 Lefkas earthquake was examined. On 17 November 2015 (07:10 GMT) a major earthquake (Mw=6.4) occurred on the central-western part of Lefkas island. Several destructive events were located in the past in this fault zone, so an extensive seismotectonic study is feasible for that area. Manual analysis was performed using a custom velocity model that was determined for that purpose, applying the average travel-time residuals and location uncertainties errors minimization method. Several clusters belonging to the aftershock sequence were identified, whereas three are directly related to the causative fault, covering an area of about 25 km. The central one, which includes the mainshock, comprises of only a few aftershocks. The northern, within which the majority of aftershocks are located, lies in the central part of Lefkas island and the southern ______

*Corresponding author: E-mail: [email protected];

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

occurred close to the SW edge of the island. In addition, offshore clusters with distinct characteristics have been identified to the south, between Lefkas and Cephalonia islands. The temporal evolution of the aftershock sequence indicates that no migration was observed, given that after the occurrence of the mainshock the entire epicentral area was activated. Focal mechanisms of the Seismological Laboratory of the University of Athens showed dextral strike-slip faulting for both mainshock and major aftershocks of the sequence. Taking into account the spatial distribution of the aftershocks, supported by the tectonic and geomorphological settings of the region, a deformation pattern, consisting of the Cephalonia-Lefkas and Ithaca-Lefkas major fault zones which converge in the area of Vassiliki bay is proposed. The appearance of the southernmost clusters was interpreted by the positive Coulomb stress changes transfer due to major earthquake Mw=6.4.

Keywords: Seismotectonic analysis; releasing bends; restraining bends; lefkas aftershock sequence; Western Greece.

1. INTRODUCTION as the Pre-Apulian (or Paxos) and Ionian [14], forming the External Hellenides terrain. These The Hellenic peninsula is bounded by important structures which have resulted by a subduction- geologic and tectonic features such as the Alpine related compressional regime are coaxial with mountain chain to the North, caused by the earlier Alpine ones resulting from the collision of collision between Europe and Africa [1], the the Pre-Apulian plate with Eurasia. The mountain North Anatolian Fault zone to the East chain of Hellenides located between southern [2,3,4,5,6], created by the lateral motion of Albania and the Gulf of Corinth plays an Anatolia with respect to the European tectonic important role on the seismotectonics of the plate [7] and the Hellenic arc to the south, region. characterized by the subduction of Tethys oceanic crust [8]. The Ionian Islands are separated from the mainland by rapidly extensional Pliocene- The area of Western Greece (Fig. 1.) which is Quaternary basins [15,16]. The oldest sediments running from the Greek-Albanian borders to the that fill those basins are of Pliocene age. In southern edge of Peloponnese and from the Lefkas Island, carbonate and clastic sediments of Ionian Islands to eastern Thessaly and the Hellenic arc external geotectonic units Macedonia lies between a continental collision dominate the geological setting [17]. The zone to the north (Adria Microplate-Eurasia) and geotectonic units of Paxos and Ionian are the Hellenic Trough to the south, where separated by a major west-directed thrust [18, remnants of the Eastern Mediterranean oceanic 19,14], marked by Triassic evaporitic domes [20]. crust subduct under the Aegean continental lithospheric microplate. These zones are linked The majority of strong earthquakes in the with the Cephalonia-Lefkas Transform Fault broader region have mainly occurred along the Zone (CLTFZ), playing an important role in the main tectonic features, such as the CLTFZ. geodynamics of the area. Seismological data for Lefkas Island is characterized by the occurrence the CLTFZ indicate right-lateral strike-slip focal of large earthquakes, both during the historical mechanisms [9,10] which is in agreement with and the instrumental era, causing significant the geodetic data revealing the NNE–SSW damage [21,22]. Most events were located close direction of slip motion [11,12]. The seismic to the NW part of the island. More specifically, strain rate is well correlated with the principal the 22 November 1704 (M=6.3), the 12 October horizontal axis of the total geodetic strain rate 1769 (M=6.7), the 23 March 1783 (M=6.7), the field. Some smaller neotectonic structures are 28 December 1869 (M=6.4) and the 27 mainly expressed by minor faults that strike November 1914 (M=6.3) earthquakes were NNE-SSW to E-W direction, overprinting the pre- among the most significant ones. They caused existing thrust-related Plio-Quaternary features, several deaths, injuries, collapse of buildings, breaking up the island in multiple independent fissures, liquefactions and landslides at the fault blocks [13]. northwestern and central parts of the island. This is the reason why most epicenters of the The broader area of study is characterized by a historical earthquakes are located close to the series of NW–SE striking geotectonic units, such northern end of the CLTFZ, in the Ionian Sea [23,

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

24]. More recently, on 14 August 2003 (05:14 (2015) earthquake sequence, in the vicinity of GMT), a large earthquake (Mw=6.3) with a focal CLTFZ. depth of 9 km occurred close to the NW coast of Lefkas Island [18,25,26]. 2.1 Construction of Local Velocity Model

On the other hand, only two large events have In the present study, the best located events been located close to the southwestern edge of were used to obtain an accurate local 1D velocity the Lefkas Island, an area that belongs to the model for the broader area of Lefkas Island. The central part of the CLTFZ. Nevertheless, initial earthquake locations for the area of important microseismic activity is observed. The Western Greece were obtained using the two events, related to this area, are the 22 regional velocity model derived by [29]. The February 1723 (M=6.7) and the 22 April 1948 selected events had at least fourteen (14) P- and (M=6.5) earthquakes. Concerning the latter eight (8) S-wave arrival times. The present model event, it caused damage at the SW part of the was determined by applying the average travel- island, while fissures and tsunami waves were time residuals and location uncertainties errors also observed. Two months later, on 30 June (RMS, ERX, ERY, ERZ) minimization method 1948, an earthquake of magnitude M=6.4 [30,31,32]. The obtained local velocity model occurred at the NW part of the island. (Table 1) yielded improved hypocentral solutions with smaller errors than those derived by the A strong earthquake of moment magnitude initial one (Table 2).

Mw=6.4 occurred on 17 November 2015 at the The value of Vp/Vs ratio was obtained using the western part of Lefkas Island, causing some following methods: a) Chatelain (1978), that damage, landslides and ground fissures. In the consists of determining the slope of the straight present study, this earthquake sequence is best-fit line of the difference between S-wave investigated in detail. For that purpose, precise and P-wave travel-times for each couple of hypocentral locations are required. The latter stations and for each event and b) the travel-time were obtained using a local velocity model that is residuals and location uncertainties errors (RMS, determined in this study. ERX, ERY, ERZ) minimization method. The last mentioned Vp/Vs ratio determination method 2. DATA AND METHODS follows the same procedure that was performed in order to define velocity and ceiling depth for The present study focuses in the area of Central each layer. The data for both Chatelain and Ionian Islands (Western Greece), where data of Spatio-Temporal Error Minimization methods several local stations belonging to the Hellenic converge to the same Vp/Vs ratio value of 1.79. Unified Seismological Network (HUSN) [27] were Error statistics for all spatial groups with both used for the construction of the local 1D velocity regional and custom local models are presented model. HUSN (Fig. 2.) comprises stations from in Table 2. The mean horizontal (ERH) and the Seismological Laboratory of the University of vertical (ERZ) location errors of the events Athens (S.L-U.O.A), the Geodynamics Institute of located in the broader study area are smaller the National Observatory of Athens (GE.IN.- than 1 km, while the mean RMS error is 0.132 N.O.A), the Geophysical Laboratory of the sec. University of Thessaloniki (G.L.-A.U.TH) and the Seismological Laboratory of the University of 2.2 Coulomb Stress Transfer Patras (U.P.S.L.). In order to investigate the possible acceleration The data set used in this study comprises of or triggering of the above-observed clusters due more than 10,000 earthquakes which were to the strong earthquake Mw=6.4, the Coulomb obtained from the seismological stations of stress changes transfer was calculated in HUSN and correspond to the time period 2012- different depths and cross-sections. The 2017 (Fig. 2). More specifically, all the event transferred ΔCFF was determined on the fault locations were calculated using manually picked plane with an effective coefficient of friction μ=0.4 P- and S-wave arrival-times, the HYPOINVERSE [33,34]. algorithm [28] a regional 1-D velocity model, 3. RESULTS AND DISCUSSION while the duration magnitude was calculated according to the formula described by relevant On 17 November 2015 (07:10 GMT) a studies [28] [10]. A subset of the data catalogue destructive earthquake occurred along the SE corresponds to the Paliki (2014) and Lefkas coast of Lefkas Island, in the vicinity of Athani

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

village. There were two fatalities, one due to 3.1 Description of Clustered Seismicity rockfall and one caused by a paddock collapse. Eight people were injured, two of which children. The hypocentral locations, using the newly Certain collapses, rockfalls and landslides have developed 1-D local velocity model, revealed that been reported. the aftershock sequence mainly occurred in the mid-western part of Lefkas Island. However, Manual analysis was conducted for the another group of events is located S of Lefkas mainshock and for more than 2,500 aftershocks Island, offshore, towards Cephalonia Island. It is for the period between 17 November and 3 worth noting that the largest aftershock (Mw=5.0) December 2015, during which the major part of occurred near the mainshock, on 17 November the ruptured area was activated. Thus, it is 08:33 UTC. Both mainshock and the largest considered that this period provides all aftershock of 17 November 2015 are well aligned necessary information to interpret the 2015 in a SSW-NNE direction. Several linear Lefkas seismic sequence. structures can be distinguished, many of which

Fig. 1. a) Main tectonic and volcanic elements of the broader area of Greece [35,29,36]. The study area is highlighted with a white rectangle b) Zoom in the study area of Central Ionian Sea, showing the islands of Cephalonia and Lefkas. Abbreviations in both maps are as following: HT, Hellenic Trench; NATr, North Aegean Trough; HVA, Hellenic Volcanic Arc; CLTFZ, Cephalonia-Lefkas Transform Fault Zone; PF, Paliki Fault; NVF, Nydri-Vassiliki Fault

Fig. 2. Distribution of HUSN stations (red triangles) throughout Greece

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

Table 1. Regional and custom velocity model

Karakonstantis (2017) VP/VS=1.79 This Study VP/VS=1.79 Layer VP (km/s) Depth (km) VP (km/s) Depth (km) 1 5.3 0.0 4.9 0.0 2 6.0 6.0 5.2 4.0 3 6.3 11.0 5.9 7.0 4 6.5 14.0 6.2 11.5 5 6.7 21.0 6.4 13.0 6 7.3 39.0 6.5 16.0 7 8.0 80.0 7.3 39.0

Table 2. Statistics of location uncertainties and median depth for the regional and custom model

Karakonstantis (2017) This Study Mean RMS (s) 0.139 0.132 Median RMS (s) 0.110 0.110 Mean ERH (km) 1.125 0.909 Median ERH (km) 0.730 0.580 Median ERX (km) 0.510 0.380 Median ERY (km) 1.070 1.050 Mean ERX (km) 0.614 0.459 Mean ERY (km) 1.636 1.360 Mean ERZ (km) 1.933 1.323 Median ERZ (km) 1.310 0.870 Mean Depth (km) 7.542 8.765 Median Depth (km) 6.860 8.375 St.Dev.Y (km) 13.036 13.076 St.Dev.X (km) 6.569 5.963 St.Dev.Z (km) 3.550 3.459

Fig. 3. Map of the study area showing the seven (7) main spatial clusters marked with different colors: cluster #1 (orange), cluster #2 (cyan), cluster #3 (blue), cluster #4 (green), cluster #5 (red), cluster #6 (yellow) and cluster #7 (purple). The back ground seismicity, before and after the main body of the aftershock sequence is marked with grey open circles

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

Fig. 4. SSW-NNE cross-section of the main body of the aftershock sequence. Top axis is marked by the values of latitude (°N) that the thin section passes by offshore, south of Lefkas island, trending roughly shallower events) down to ~15 km and includes E-W. The aftershock distribution appears mostly the second largest aftershock (Mw=5.0) which aligned in a N20°E direction with several occurred on 18 November 2015, 12:15 UTC. The branches oriented ~N30°E, including the cluster could be further divided in 2 sub-clusters, northernmost and southernmost tips (Fig. 3). the one for the shallower, roughly onshore events, trending S-N and the other, offshore The cross-section drawn at N20°E (Fig. 4.), further north, containing fewer events and roughly parallel to the CLTFZ, indicates that the apparently trending SW-NE. total length of the activated area is approximately 60 km. The focal depths are distributed between At the southern group, the northern tip of cluster 5 and 15 km, with the clusters in the northern #4 is ~4-5 km south of the mainshock and the group (5 - 7) being generally shallower than cluster extends about 8 km SSW while its those of the southern group (1 - 4). The median focal depth (~12 km) is larger than the mainshock, as well as the strongest aftershock one for cluster #5 (~9 km). It probably belongs to (Mw=5.0) of 17 November are contained in the same fault plane as the one of the main cluster #5. rupture and defines its deeper seismogenic part. Clusters #2 and #3 are less dispersed than The epicenter of the major events of this cluster cluster #4 and their distribution is roughly are mainly aligned SSW-NNE, with the smaller oriented E-W. Cluster #2, in particular, contains 2 sub-cluster being apparently oriented E-W near large sub-clusters at 8 – 13 km depth and a Vassiliki bay. By far the largest cluster is #6, smaller sub-cluster at 15 km depth. The which contains more than 100 events of the southernmost cluster #1 is a bit offset from the aftershock sequence (17/11/2015-03/12/2015). cross-section line (Fig. 4). Its main body is Cluster #6 is divided into two branches in its located at 7-9 km depth and it also includes shallower parts (4-6 km), one SSW-NNE, similar several sub-clusters dispersed further south. At to cluster #5, and another trending WSW towards the very southern edge of the study area, a the western coast, extending a bit deeper (8-9 group of 3 deeper events at ~23-24 km, with km) ENE. The described hypocentral distribution epicenters between Ithaki and Cephalonia is probably related to fault network complexity, islands, are also included in cluster #1. which has possibly acted as a barrier, prohibiting the main rupture to extend further north. It also The Lefkas 2015 aftershock sequence is typical contains most of the major aftershocks (13 in terms of its spatio-temporal characteristics. events with Mw≥3.9) which are, on average, Aftershocks were generated at both northern and located at slightly larger focal depths (~10 km) southern edges of the zone within minutes to than the smaller ones (~6 km). Further north, hours following the mainshock, as it is cluster #7 extends ~17 km and deepens from ~6 evident from S.L-U.o.A catalogue km onshore (excluding a few sparsely located (http://www.geophysics.geol.uoa.gr/stations/gma

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

ps3/leaf_significant.php?mapmode=mech&lng=e to the positive Coulomb stress transfer during the n&year=2015#mapanc) while the largest one fault rupture which occurred on the central- (Mw=5.0) occurred in less than 2 hours after the western part of Lefkada island, on 17 November main event, in the same cluster. By the end of 2015, (Mw=6.4). It is worth noting that the same December 2015, the activity in most of the spatial procedure was observed for the 2003 Lefkas groups had diminished (Fig. 5), with the rupture process [26]. exception of cluster #6, and the sequence was typically over, as confirmed by routine 3.4 Discussion observations of the seismicity rate during the following months. The 2015 Lefkas aftershock sequence is characterized by the existence of several 3.2 Focal Mechanisms clusters. However, the spatial distribution of the sequence revealed the activation of a large area Fault plane solutions of the largest aftershocks that extends further than the mainshock’s rupture (Mw≥5.0), determined by the Seismological length. Its first characteristic is that only few Laboratory of the University of Athens (S.L- aftershocks are located in the vicinity of the U.o.A; http://www.geophysics.geol.uoa.gr ), epicenter. On the contrary the most important indicate dextral strike-slip faulting. As an cluster is situated at the central part of the island. example, the focal mechanism of the largest It is characterized by high concentration of (Mw=5.0) aftershock, occurred on 17 November events, constituting the main cluster, and it is 2015 (08:33 GMT), 4 km SSW of the mainshock, observed towards the northern end of the is presented in Fig. 5. activated fault segment, at the mid-northern part of Lefkas Island. In the same area, an important Focal mechanisms revealed the activation of an cluster occurred during the 1994 sequence, as almost vertical right lateral strike slip fault, in described by [21]. More specifically, on 29 agreement with the ENE-WSW oriented CLTFZ. November 1994 a moderate event of Mw=5.1 The obtained results indicate that the dimensions took place on the northern part of CLTFZ, close of the activated fault, located between the main to the west coast of Lefkas Island, followed by a aftershock cluster in the central part of the island M=4.8 aftershock on 1 December 1994. and the respective one SSW of Lefkas, are 25 Aftershock activity was concentrated in the west- km length and 10 km width. The focal central part of the island, in the region where the mechanisms of the events located offshore, main cluster of the 2015 sequence occurred. In between Lefkas and Cephalonia Islands, are addition, the 1994 sequence was characterized similar to the respective solution of the by an almost vertical distribution, reaching 12 km mainshock. Nevertheless, taking into account depth [21]. The above-mentioned observations additional characteristics, such as the spatial suggest that this area produces a complex distribution of the aftershocks and the shallower seismicity pattern, which is probably due to the bathymetry, both planes of the focal mechanism existence of a local minor fault system. This is of the offshore events could be considered as the supported by the high seismic rates, observed in activated fault, as it will be discussed in more the central part of Lefkas Island following the detail in the following section. occurrence of moderate or large local events.

In addition, spatially clustered seismic activity 3.3 Coulomb Stress Transfer results occurred to the south, close to the NNW part of Cephalonia Island, probably is not directly related The ΔCFF distribution, as well as the cross to the main activated fault since they are not section parallel to the fault, is presented in figure located along the CLTFZ but they are shifted 6a and 6b respectively. The Coulomb stress eastwards forming a step over of the ruptured distribution showed that the earthquake with zone. The southern clusters (#1, #2 and #3) are magnitude Mw=6.4 caused strong positive stress stretched in a roughly WNW-ESE direction, changes transfer in the directions NNW and separated by gaps between them, possibly SSE, (Fig. 6c.) as well as in the upper layers indicating the existence of small parallel left- from the depth of 8Km until the surface (Fig. 6d.). lateral structures, transverse to the main CLTFZ. Additionally, the max stress coulomb values were Another couple of small clusters of undefined calculated for a depth range 0-20Km with step geometry are located near the southern coast of 2Km (Fig. 6e.). The results lead to the conclusion Lefkas Island (spatial group #4). Other than that, that the southernmost clusters are triggered due the vicinity of the mainshock’s hypocenter is

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

characterized by few sparse aftershocks. The distributed in the northern segment of the studied area defined by the spatial groups #5 and fault zone, some significant spatial clusters to the possibly #4 is considered to be the main rupture south of Lefkas Island appeared almost area, likely a barrier that left a few unbroken perpendicular, in a mean WNW-ESE direction, asperities where the seismic clusters are with the focal mechanisms implying sinistral observed. However, certain clusters occurred to strike-slip faulting. The results of the present the south, between Lefkas and Cephalonia study are in general fairly correlated to the Islands. Taking them into account, the spatial scenario of previous seismological studies in the distribution spans roughly ~60 km in a SSW-NNE area [20], with the exception of the smaller than direction, while focal depths vary between 5 km expected, according to their model, fault’s (mostly onshore) and 15 km (offshore). The dimensions, taking into account the southern temporal evolution of the aftershock sequence is clusters of the 2015 Lefkas sequence. Thus, the generally smooth. No significant migration perpendicular secondary fault system of the latter patterns were observed, as almost the full extent model has to be prolonged towards the north, up of the aftershock zone was activated within the to the southern part of Lefkas. In that case, the first 24 hours after the mainshock. negative flower structure model, proposed by [20], could be adjusted by making a double left- Recent seismotectonic studies suggest that bend which explains lots of the superficial either the CLTFZ in this area is collinear with its geological and geomorphological features at the northern segment west of Lefkas, while further SW edge of Lefkas [13,37,7,38]. The advantage south it is shifted by some kilometers to the west of this model is that it explains the observed of Cephalonia coast due to a transfer zone of seismicity south of Lefkas, while its disadvantage extensional step overs [27,11] or that the CLTFZ is the steep bending of the main part of the fault does not consist of a single dextral strike-slip zone which is not compatible with the rupture fault, but rather a system of faults. Even though process of the mainshock (Fig. 7a.). the 2015 Lefkas aftershock sequence was mainly

Fig. 5. Evolution of the seismicity in the study area before (2012-2015a) and after (2015b-2017) the occurrence of November 17, 2015 M6.4 Lefkas earthquake

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

Fig. 6. Distribution maps of Coulomb stress changes (ΔCFF) due to the 2015 earthquake Mw=6.4 at the focal depth (a) and cross section AB for strike 10°N (b). ΔCFF Maps for depths 1Km (c), 7Km (d), respectively. Max values of Coulomb stress changes for depth range 0-20Km (e)

Fig. 7. a) Double left-bend of CLTFZ model (scenario 1) and b) two major converging fault zones (CLTFZ and ILFZ), forming an almost antithetic system of strike-slip faults (scenario 2)

Another scenario, which fits most of the results elements at the southern part of Lefkas and the obtained by this study, is combined with previous sea-bottom of Myrtos Gulf [13,7,38]. It includes knowledge based on GPS measurements [39], two major fault zones, the CLTFZ and Ithaca- geophysical [40] geomorphological and tectonic Lefkas (ILFZ), converging in the area of Vassiliki

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

bay forming en-enchelon arrays of WNW-ESE that was also revealed for the two earthquakes minor faults and fractures, separated by smaller that occurred in Cephalonia during January – NE-SW strike-slip faults in between (Fig. 7b), February 2014 [10]. The events located close to such as the one of Nydri-Vassiliki (NVF). In this Lefkas Island are related to the main tectonic model, there is the major NNE-SSW striking feature of the area, which is the Cephalonia- CLTFZ offshore, which makes a right bend NW Lefkas Transform Fault Zone, along which other of Myrtos bay, forming a releasing bend in a significant earthquakes occurred in the past, region SSW of Lefkas Island, as in California’s such as the 1983 Cephalonia earthquake. deformation model described by [41]. On the According to the results of the present study, the other hand, the NNW-SSE ILFZ and CLTFZ ruptured fault, with a length of approximately 25 converge in western Lefkas Island, forming an km, is the prolongation of the 2003 Lefkas antithetic system of strike-slip faults to the overall earthquake causative fault. sense of the zone’s shear, comprising crustal rotated blocks of oblique minor faults in the area The occurrence of the 2015 Lefkas earthquake between Cephalonia and Lefkas Islands sequence provided the opportunity to obtain a (Fig. 7b.). more detailed deformation pattern, especially for the area between Lefkas and Cephalonia 5. CONCLUSION islands. For that purpose, two major scenarios are investigated. The first is a negative flower The 17 November 2015 Mw=6.4 Lefkas structure model prolonged towards the southern earthquake is located at the southwestern part of part of Lefkas. The other scenario includes the the island and can be considered as the Cephalonia-Lefkas and the Ithaca-Lefkas major continuation of the 2003 ruptured area, which fault zones, converging in the area of Vassiliki probably left an unbroken patch. It is obvious that bay, separated by smaller strike-slip faults in the recent event occurred within this asperity. between, almost perpendicular to the CLTFZ. Taking into account both events, the activated Considering the second scenario, both the area, with a length of about 45 km in a SSW- WNW-ESE clustered seismicity, between the NNE direction, covers the northern part of CLTFZ islands of Cephalonia and Lefkas, and the uplift that is located west of Lefkas. This observation region in the SW part of Lefkas Island, can be implies that the accumulated energy was reasonably explained. released in two separate events instead of one, which would obviously have considerably larger COMPETING INTERESTS magnitude. Such earthquake sequences, consisting of more than one event, have been Authors have declared that no competing reported since the historical times, as in the case interests exist. of the earthquakes that occurred during 1948 [21]. REFERENCES

The main characteristic of the sequence is that 1. Anderson H, Jackson J. Active tectonics of only a few aftershocks with M≥3.0 are located the Adriatic Region. Geophysical Journal close to the mainshock’s epicentre, suggesting of the Royal Astronomical Society, 1987; that during the rupture process, no unbroken 91:937-983. patches were left at the central part of the DOI:10.1111/j.1365-246X.1987.tb01675.x activated fault. The southern part of the activated fault, reaching the SW edge of Lefkas Island, is 2. Cisternas A, Polat O, Rivera L. The characterized by low seismicity. Concluding, the Marmara Sea Region: Seismic behaviour total length of the causative fault is of the order of in time and the likelihood of another large 25 km, while the seismicity belonging to the earthquake near Istanbul (Turkey), Journal southern and northernmost (1-3 and 7) clusters of Seismology. 2004;8:427-437. is probably not directly related to the CLTFZ. The 3. Gurbuz C, Aktar M, Eyidogan H, Cisternas length of the total activated area is about 60 km, A, Haessler H, Barka A, Ergin M, Turkelli much larger than the expected rupture length for N, Polat O, Ucer SB. et al. On a Mw=6.4 mainshock. This observation suggests seismotectonics of the Marmara region that the clusters #1, #2, #3 and #7 have been (Turkey): Results from a microseismic triggered by stress-transfer. experiment, Tectonophysics. 2000;316:1- 17. The majority of the determined focal mechanisms 4. Polat O, Haessler H, Cisternas A, Philip H, are similar, related to dextral strike-slip faulting Eyidogan H, Aktar M, Frogneux M, Comte

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

D, Gurbuz C, The Izmit (Kocaeli) Turkish western Greece. Tectonophysics. 1984; earthquake of August 17, 1999: Previous 101:25–54. seismicity, aftershocks and 16. Zelilidis A, Kontopoulos N, Piper DJW, Avramidis P. Tectonic and seismotectonics, Bulletin Seismological sedimentological evolution of the Pliocene- Society of America. 2002b;92(1):361-375. Quaternary basins of Zakynthos Island, 5. Polat O, Haessler H, Cisternas A, Philip H, Greece: Case study of the transition from Eyidogan H, Analysis and interpretation of compressional to extensional tectonics. the aftershock sequence of the August 17, Basin Research. 1998;10:393-408. 1999, Izmit (Turkey) earthquake, Journal of 17. Renz C. Die vorneogene stratigraphie der Seismology. 2002a;6:287-306. normalsedimentaren Formationen 6. Polat O, Gok E, Yilmaz D. Earthquake Griechenlands, Greece. Inst. Geol. Hazard of Aegean Extension Region, Subsurf. Res. Athenes. 1955;1-637. Turkey, Turkish Journal of Earth Sciences. 18. Benetatos C, Kiratzi A, Roumelioti Z, 2008;17:593-614. Stavrakakis G, Drakatos G, Latoussakis I. 7. Bathrellos G, Antoniou V, Skilodimou H. The 14 August 2003 Lefkas Island Morphotectonic characteristics of Lefkas (Greece) earthquake: Focal mechanisms Island during the Quaternary (Ionian Sea, of the mainshock and of the aftershock Greece). Geologica Balcanica. 2009;38: sequence. Journal of Seismology. 2005; 23-33. 9(2):171-190. 8. van Hinsbergen DJJ, van der Meer DG, 19. Bornovas J. Géologie de l’île de Lefkade. Zachariasse WJ, et al. Int J Earth Sci (Geol Geol. and Geophys. Research (I.G.S.R.). Rundsch). 2006;95:463. 1964;10:142. Available:https://doi.org/10.1007/s00531- 20. Karakitsios V, Rigakis N. Evolution and 005-0047-5. petroleum potential of Western Greece. J. 9. Scordilis EM, Karakaisis GF, Karacostas Pet. Geol. 2007;30:197–218. BG, Panagiotopoulos DG, Comninakis PE, 21. Makropoulos K, Kaviris G, Kouskouna V. Papazachos BC. Evidence for transform An updated and extended earthquake faulting in the Ionian Sea: the Cephalonia catalogue for Greece and adjacent areas island earthquake sequence of 1983. since 1900, Nat. Hazards Earth Syst. Sci. PAGEOPH. 1985;123:387–397. 2012;12:1425-1430. 10. Papadimitriou P, Chousianitis K, Agalos A, 22. Stucchi M, Rovida A, Gomez Capera AA, Moshou A, Lagios E, Makropoulos K. The Alexandre P, Camelbeeck T, Demircioglu spatially extended 2006 April Zakynthos MB, et al. The SHARE European (Ionian Islands, Greece) seismic sequence Earthquake Catalog (SHEEC) 1000–1899. and evidence for stress transfer. Geophys. J. Seismolog. 2013;17(2):523–544. Jour. Intern. 2012;190(2):1025-1040. 23. Kouskouna V, Makropoulos KC, Tsiknakis 11. Kahle HG, Muller MV, Veis G. Trajectories K. Contribution of historical information to a of crustal deformation of western Greece realistic seismicity and hazard assessment from GPS observations 1989–1994. of an area. The Ionian Islands earthquakes Geophys. Res. Lett. 1996;23(6):677–680. of 1767 and 1769: Historical investigation. 12. Lagios E, Sakkas V, Papadimitriou P, In: Stucchi, M. (Ed.), “Historical Damiata BN, Parcharidis I, Chousianitis K, Investigation of European Earthquakes,” et al. Crustal deformation in the Central Materials of the CEC project “Review of Ionian Islands (Greece): results from Historical Seismicity in Europe”. 1993;1: DGPS and DInSAR analyses (1995-2006). 195–206. Tectonophysics. 2007;444:119–145. 24. Makropoulos KC, Kouskouna V. The 13. Lekkas E, Danamos G, Lozios S. Ionian Islands earthquakes of 1767 and Neotectonic structure of Lefkas Island, 1769: seismological aspects. Contribution Bull. Geol. Soc. Greece, XXXIV. 2001;1: of historical information to a realistic 157-163. seismicity and hazard assessment of an 14. Jacobshagen VH. Geologie von area. In: Albini, P., Moroni, A. (Eds.). Griechenland. Beiträge zur regionalen “Historical Investigation of European Geologie der Erde 19. Berlin; 1986. Earthquakes,” Materials of the CEC Project 15. Brooks M, Ferentinos G. Tectonics and “Review of Historical Seismicity in Europe”. sedimentation in the Gulf of Corinth and 1994;2:27–36. the Zakynthos and Kefallinia channels,

Karakonstantis et al.; JGEESI, 21(1): 1-12, 2019; Article no.JGEESI.49088

25. Zahradnik J, Serpetsidaki A, Sokos E, 34. Toda S, Stein R, Lin J, Sevilgen V. Tselentis GA. Iterative Deconvolution of Coulomb 3.2, Graphic-rich deformation & Regional Waveforms and a Double-Event stress-change software. User Guide; 2010. Interpretation of the 2003 Lefkas 35. Ganas A, Oikonomou I, Tsimi C. Earthquake, Greece. Bull. Seism. Soc. Am. NOAfaults: A digital database for active 2005;95(1):159–172. faults in Greece. Bulletin of the Geological 26. Papadimitriou P, Kaviris G, Makropoulos Society of Greece. 2013;47(2):518-530. K. The Mw = 6.3 2003 Lefkas Earthquake (Greece) and induced stress transfer DOI:http://dx.doi.org/10.12681/bgsg.11079 changes, Tectonophysics. 2006;423:73– 36. Pavlides S, Caputo R, Sboras S, 82. Chatzipetros A, Papathanasiou G, 27. D’Alessandro A, Papanastassiou D. Valkaniotis S. The Greek catalogue of Baskoutas I. Hellenic Unified active faults and database of seismogenic Seismological Network: an evaluation of its sources, Bulletin of the Geological Society performance through SNES method, of Greece. 2010;XLIII(1):486-494. Geophys. J. Int. 2011;185:1417–1430. 37. Margaris B, Papaioannou C, Theodulidis 28. Kaviris G, Papadimitriou P, Makropoulos N, Savvaidis A, Anastasiadis A, Klimis N, K. Magnitude Scales in Central Greece. et al. “Preliminary Observations on the Bull. Geol. Soc. Greece. 2007;XXXX(3): August 14, 2003 Lefkas Island (Western 1114-1124. Greece) Earthquake”, EERI Special 29. Karakonstantis A. 3-D simulation of crust Earthquake Report, November, (Joint and upper mantle structure in the broader report by Institute of Engineering Hellenic area through Seismic Seismology and Earthquake Engineering, Tomography. Dissertation, National and National Technical University of Athens & Kapodistrian University of Athens; 2017. University of Athens. 2003;1-12. Greek 38. Triantaphyllou MV. Calcareous nannofossil 30. Kissling E, Ellworth WL, Eberhart-Phillips biostratigraphy of Langhian deposits in D, Kradolfer U. Initial reference models in Lefkas (Ionian Islands). Bulletin of the Geological Society of Greece. Proceedings local earthquake tomography, Journal of th Geophysical Research. 1994;99:19635− of the 12 International Congress. 2010; 19646. XLII(2):754-762. 31. Chiarabba C, Frepoli A. Minimum 1D 39. Sakkas V, Lagios E. Fault modelling of the velocity models in Central and Southern early-2014 ~M6 Earthquakes in Italy: a contribution to better constrain Cephalonia Island (W. Greece) based on hypocentral determinations. Ann. GPS measurements. Tectonophysics. Geophys. 1997;40(4):937-954. 2015;644–645:184–196. 32. Lopes AEV, Assumpção M. Genetic Available:http://dx.doi.org/10.1016/j.tecto.2 Algorithm Inversion of the Average 1D 015.01.010. Crustal Structure using Local and Regional 40. Kokinou E, Kamberis E, Vafidis A, Earthquakes. Computers & Geosciences; Monopolis D, Ananiadis G, Zelilidis A. 2011 Deep seismic reflection data from offshore DOI:10.1016/j.cageo.2010.11.006. western Greece: A new crustal model for 33. Toda S, Stein RS, Richards-Dinger K, the Ionian Sea, Journal of Petroleum Bozkurt S. Forecasting the evolution of Geology. 2005;28(2):185-202. seismicity in southern California: 41. Twiss RJ, Moores EM. Structural Geology. Animations built on earthquake stress W.H. Freeman and Company, New York; transfer. J. Geophys. Res. 2005;B05S16, 1992. DOI:10.1029/2004JB003415. ______© 2019 Karakonstantis et al.; This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Peer-review history: The peer review history for this paper can be accessed here: http://www.sdiarticle3.com/review-history/49088